Google Antitrust Remedies: Promoting Competition Without Punishing Users

Considering how to increase competition in the search space without damaging end users is a trickier question than it seems at first. Many of the suggestions that people have tossed out have tended to focus on ideas that are purely punitive to Google, but which would also have negative impacts on users (and even some competitors). As we reach the stage of the antitrust battle where remedies are actually being considered, it’s crucial that we focus on solutions that will truly promote competition and benefit users, not just score political points against Google.

Earlier this year, I was left troubled by the end result of the ruling against Google in the first (of a few) antitrust cases against it. I think the (currently ongoing) case about the company’s practices regarding advertising is a lot stronger. The case that was ruled on this summer, though, was about Google’s massive payments to Apple and Mozilla to have those companies have Google search as the default on Apple devices/Safari and on Firefox.

At the time, we pointed out that it was difficult to think of any remedies that actually helped solve the situation. Both Apple and Mozilla more or less admitted during the trial that users effectively demanded Google search be the default, and any attempt to use other search engines resulted in angry users. If the court demanded Google stop paying the billions of dollars to Apple or the hundreds of millions of dollars to Mozilla, it wouldn’t hurt Google. Indeed, it would seem to help them.

Since both companies admitted that users were demanding Google as the default, little would change there other than Google getting to keep even more money. And Apple and (especially) Mozilla losing a ton of revenue. That didn’t seem very helpful at all.

In the intervening months, I’ve had a few conversations with folks about possible remedies that make sense. The most reasonable suggestion seemed to be DuckDuckGo’s main suggestion: allow other search engines to build off of Google’s search corpus by enabling API access under Fair, Reasonable and Non-Discriminatory (FRAND) grounds.

The best and fastest way to level this playing field is for Google to provide access to its search results via real-time APIs (Application Programming Interfaces) on fair, reasonable, and non-discriminatory (FRAND) terms. That means for any query that could go in a search engine, a competitor would have access to the same search results: everything that Google would serve on their own search results page in response to that query. If Google is forced to license its search results in this manner, this would allow existing search engines and potential market entrants to build on top of Google’s various modules and indexes and offer consumers more competitive and innovative alternatives. 

Today, we believe that we already offer a compelling search alternative with more privacy and fewer ads, relative to Google. We’ve also been working for fifteen years to make our search results on par in terms of feature set and quality by combining our own search indexes with those of partners like Apple, Microsoft, TripAdvisor, Wikipedia, and Yelp. However, we know that many consumers still prefer Google’s results due to the benefits of scale discussed above, and this intervention would erase that advantage, instantly making us and others much more competitive. 

This remedy would certainly allow for more competition to arise, which has proven difficult today. No one (not even Microsoft’s Bing) really has the reach and comprehensiveness of Google’s index. DuckDuckGo is mostly built on Bing (I know it insists it’s more than that, but in practice, it appears to be mostly Bing — as we discovered when Bing banned Techdirt, and we also disappeared from DDG).

Every attempt to build competing search engines seems to run into the scale problem eventually without access to Google results. Even Kagi, which was briefly a darling among folks looking for a search alternative, apparently makes use of Google’s search tech on the backend. It seems like a pretty reasonable idea to make it so that others can license access to the API and build Google results into alternative search products, as this gets at the actual issues underlying this case.

A few weeks ago, the Justice Department filed its preliminary thoughts on remedies, and there are a wide mix of ideas in there, some crazier than others. A lot of the headlines that filing generated were around big “break up” ideas: spinning off Chrome or Android. These seem preposterous and unlikely. Under antitrust law while breakups (“structural remedies”) are certainly one tool in the toolbox, they are supposed to be related to the violation at hand.

Given that the antitrust problem in this case was about the search payments, and not anything specific to Chrome or Android, it’s difficult to see how such remedies would even be allowed under the law, let alone make sense. Indeed, without Chrome and Android being attached to Google, those products would likely suffer, as both are subsidized by Google, and that would do a lot to harm users. That doesn’t seem like a good result either.

So the proposals from the DOJ that match DDG’s suggestion of API access are much more interesting (and probably better) overall.

Plaintiffs are considering remedies that will offset this advantage and strengthen competition by requiring, among other things, Google to make available, in whole or through an API, (1) the indexes, data, feeds, and models used for Google search, including those used in AI-assisted search features, and (2) Google search results, features, and ads, including the underlying ranking signals, especially on mobile

Again, this seems to actually target the issue. It creates a scenario for increased competition without a corresponding harm to users or to other competitors. Many of the other sections do not.

Also, arguably, the DOJ could have gone even further, conveying on users more ability to designate access to information and data as a way to escape the silo of Google. This is a bigger issue and one that doesn’t get as much attention, but the ability of large companies to lock in users has diminished the ability of competitors to grow and challenge the network effects of existing businesses.

For some users of Google, the fact that it tracks your history is not seen as creepy or privacy invading, but rather a benefit for that user (and yes, this is not true for everyone!). But if the user could retain control over their own search histories and preferences, and allow third party search engines to access it with the user’s permission it would also help users get out of an existing silo.

Just as one example, Google knows a fair bit about what I normally search on and click on. But if I could make use of that history and give DuckDuckGo or Kagi or someone else access to it for the sake of improving their own search results to my queries, that would be potentially useful for competition. And all it’s really doing is saying that the user who generated that history and metadata should have some control over it as well, including separating it from the underlying Google product.

Yes, this would have to be done carefully, to avoid (say) exposing more sensitive data regarding searches to these other companies, but if it was done in a way that was transparent, and which the end user had control over, it could be really valuable.

Not surprisingly, Google is very, very upset about all these potential remedies. It suggests that if they were forced to share such things with others, it would lead to privacy and security risks:

Forcing Google to share your search queries, clicks, and results with competitors risks your privacy and security. It’s widely recognized, including explicitly by the DOJ in its outline, that forcing the sharing of your searches with other companies could create major privacy and security risks. The search queries you share with Google are often sensitive and personal and are protected by Google’s strict security standards; in the hands of a different company without strong security practices, bad actors could access them to identify you and your search history — as we’ve seen before. Additionally, while sharing Google’s search results with others might create a few copycats, it could also decrease incentives for other companies to actually innovate in search.

This very much depends on what information is shared, with whom, and how. I still think that simply giving the user more control over it, rather than just letting companies fight over access, solves some of Google’s stated concerns.

On the whole, the larger structural remedies (spinning off lines of business) don’t seem to target the underlying issue, seem mainly punitive, and won’t do much to help competition or users. But the idea of opening up access to search systems and data, especially if it gives more control to the end user actually seems like a really good way of increasing competition and improving the situations for users.

Google’s statements about security and privacy are still ones worth considering, but there are ways to deal with those issues, mainly by providing more power to the end user, rather than just opening up that info directly to other search engines.